Warmboard Design Issues: Flow, Temp, and Air Removal

REKBDR

I’ve got several questions about the design of a radiant floor system for a kitchen remodeling job. I’m puzzled by issues involving flow rates when trying to size a circulating pump for the radiant loop.

THE PROJECT: The kitchen will have 168 square feet of Warmboard under oak strip flooring (R value about .6 and maximum floor temperature of 80 degrees). Warmboard’s heat chart indicates that under these conditions the output will be 3,360 BTU/hr when running an average water temperature of about 92 degrees through the 1/2” Pex-Al-Pex which is required for their system. Warmboard suggests that I shoot for .5 gpm as the flow rate. However, I am tapping into a conventional high temperature boiler system, so I will need a mixing valve (e.g., Caleffi 521) to cool the supply to the radiant loop. Moreover, the manifold (or an equivalent set of fittings, including an air separator) will be in the basement, below the elevation of the tubing.

QUESTION 1: I frequently have to bleed air out of the radiators in my old house (even though I also have an air eliminator near the boiler), and I am concerned about air collecting in the radiant loop, not only when filling but over the course of operation. Taco’s guide for selecting a pump (Tech Doc TD10) recommends a piping flow rate that is fast enough to carry along air bubbles to a separator (including one at a lower elevation) but not so fast as to be noisy. By convention this appears to in the range of 2-4 fps, which Taco converts into 1.2 to 2.5 gpm for 1/2” Pex-Al-Pex. So, I figured that I would design my single loop system for 2 gpm. However, when shopping for a manifold, I note that all the built-in flow meters for the individual loops max out at significantly less than 2 gpm. What explains the discrepancy between the flow that is typically recommended and what everybody seems to be installing? Is removing air from the individual loops not a concern for some reason, especially since the manifolds and separators are often located above the level of the tubing?

QUESTION 2: If the average water temperature in my tubing is supposed to be around 92 degrees, I want to be conservative about my design for Delta T so as not to cook my oak flooring over the initial stretches of my pex run. Hence, I have been planning for a 10 degree drop. However, even that computes a flow rate of just .66 gpm for my system (flow equals 3,360 BTU divided by 500 times 10 degrees). To achieve a 2 gpm flow rate, my Delta T would have to be only about 3 degrees. Is such a very low Delta T feasible?

QUESTION 3: Since my radiant system is just a small single loop, I am concerned about the operating parameters of the mixing valve. Typically, the valve would feed a multi-loop manifold, where the aggregate flow would be at least several gpm. In my case, the flow though the valve will be the same as the flow through my tubing. At 2 gpm this would not be a problem, but for the Caleffi 521 for example, the minimum flow for stable mixing is 1-1.3 gpm. Moreover, I have to be mindful of the specified temperature ranges of the hot and cold feeds for the valve, if I need to work with a Delta T that brackets an average temperature of about 92 degrees.. The Caleffi 521 has a cold inlet range specification that is only 39-80 degrees, which would barely work if my Delta T were at least 20 degrees (which I want to avoid). Can anyone recommend a mixing valve that would fit my needs in terms of flow and temperature ranges? Should I be looking at valves that are designed primarily as anti-scalding devices, which seem to have lower flow requirements while the lower scalding limit (120 degrees) would still be acceptable?

TWO LOOPS OR OTHER OPTIONS? How about other approaches? Since I would have an unused “extra loop” in a two loop manifold anyhow, would it make sense to divide my pex run into two loops that are even smaller? This might overcome the problem of minimum flow through the mixing valve. It would also cut the head pressure roughly in half (allowing me to use a much less expensive circulator) and help distribute the warmer and colder areas across the floor. Would shorter runs also help with the air removal issue to any significant degree?

Thanks for your help!

SV9_9

Hi, I am not qualified to answer your other questions, but you might look at the Honeywell AM100R heat only mixing valves. Spec. at .5 min GPM, and only 3 deg. min. delta T between hot and mix. Supply House stocks them.

SV9_9

They are a bit expensive but the Uponor TruFlow manifolds have a higher loop cv ( 1.9 ) than most of the others I have looked at. They also have 2 Gpm flow meters available as an accessory. I am using one of these to supply 5 large CI radiators with no apparent issues.

REKBDR

Thanks, SV9_9. The Honeywell valve looks promising. I plan to call them tomorrow to make sure that there is no problem with a limited range for the “cold” input.

As for the Uponor manifold and meter, they are pretty pricey. I have discovered that Caleffi makes meters for their manifolds in the range of up to 2 gpm and also 1-4 gpm, but their smallest manifold is three loops. I need only one loop (or maybe two).

I'm now wondering if I should just cobble together my own “manifold” and accessories. One challenge is the peculiar threads used on the different connectors.

SV9_9

You could just sweat a couple of tees together to make a simple manifold, but you will still need balancing valves and flow meters. I think you will have better luck with an off the shelf manifold. Having an extra loop isn't necessarily a problem and may prove useful in the future. If you use PAP compression fittings for the loop ends you won't need any special tools.

Hopefully you will get some response from other folks regarding your flow rate questions. It is my understanding that the 2' to 4' per sec. convention is to keep air entrained in vertical flow. It may not be as important in a small diameter horizontal tubing run.